Christmas Rushed: The game, highly anticipated before its launch, suffered numerous delays (including one cause by switching developers) before LucasArts finally put their foot down and demanded that Sony Online Entertainment release the game in summer of 2003. Unfortunately, the game was nowhere near ready and the lead developer, Raph Koster, later described frantically removing reams of incomplete code, to the point where the only incomplete code he wasn't removing were things that the game actually needed to run. With the benefit of hindsight, the end result of this rush could charitably be described as disastrous. The game was bug-ridden and largely devoid of content on launch, numerous promised features - like space travel, player-owned towns, and vehicles - had to be delayed to later patches and most of the underlying systems of the game were horrendously imbalanced and broken. Despite the best efforts of the developers, the game never truly recovered from the problems created by its early launch and most of the game's subsequent problems had their roots in the decision to gun for a Summer '03 launch window.
Apparently not learning their lesson, SOE wound up doing this a second time in 2005 with the New Game Enhancements (NGE), a particularly reviled update that completely rewrote the game code from the ground up and dramatically changed the way the game was designed and played. A long-awaited Combat Upgrade launched in May of that year had gone over poorly with the community and the developers decided they needed something bigger to draw in more players. A conceptual mock-up of what would become the NGE was created and shared amongst the developers and producers. SOE gave the project a tentative go-ahead and the devs were given a mere six months to put together a complete rewrite of all of the game's core systems, all while putting together the game's third expansion pack AND maintaining the current code with regular updates. Several of the developers would later claim that those six months involved some of the highest workload they'd ever had to deal with and because SOE still hadn't fully decided to use the new system, they had to effectively code for two different game systems at once. The NGE was finally launched with two weeks warning to the players (only two weeks because that's exactly the point when SOE confirmed they were going ahead with the switch) and managed to turn the game into an even buggier mess than it was at launch, resulting in an enormous exodus of players.
Executive Meddling: Possibly why the game was shut down. SOE said they were interested in keeping Galaxies going, but allegedly EA demanded exclusivity on Star Wars MMO's from LucasArts. Despite this, Clone Wars Adventures (a browser-based Star Wars MMO that SOE also maintained) continued to operate concurrently with TOR for almost three years.
Troubled Production: SWG was a poster child for this trope more or less throughout its entire lifespan from initial development through shutdown.
What Could Have Been: Raph Koster, the game's original creative director, has posted a series of articles on his blog detailing his original vision for Galaxies. This included a more in-depth skill tree system, various alterations to combat, and - most notably - a much different system for unlocking Jedi that rewarded players that sought out a broad-range of in-game experiences rather than promoting power-gaming.
Despite the (justified) complaints from the fanbase, the Combat Upgrade was arguably the game's highest point from a design perspective. (Most) everything worked right, professions had clearly defined roles in combat and abilities to help them get the most out of it, groups were viable and had a much easier time handling difficult content (since now everyone was damaging the same "health" bar), and even some releases of new content that kept the (remaining) players engaged. Had it been given more than six months before being replaced with the New Game Enhancements, SWG would, at the very least, be remembered far differently than it is today.